ISPRS Open Journal of Photogrammetry and Remote Sensing最新文献

{"title":"A data-driven morphological filtering algorithm for digital terrain model generation from airborne LiDAR data","authors":"Bingxiao Wu ,&nbsp;Xingxing Zhou ,&nbsp;Junhong Zhao ,&nbsp;Wuming Zhang ,&nbsp;Guang Zheng","doi":"10.1016/j.ophoto.2025.100102","DOIUrl":"10.1016/j.ophoto.2025.100102","url":null,"abstract":"<div><div>Ground filtering algorithms (GFs) are widely used in point cloud processing to generate digital terrain models. Existing GFs typically rely on rule-based or machine learning approaches to separate ground and non-ground points within an airborne point cloud. However, they often struggle to accurately extract ground points in scenarios containing mountains and heterogeneous buildings. To enhance the accuracy and robustness of ground filtering for airborne point clouds, we propose a data-driven morphological filtering algorithm (DMF). DMF begins by identifying near-ground voxel centroids after voxelizing the input point clouds. Next, a digital elevation model is constructed based on the elevation information of these near-ground voxel centroids. A composite morphological filter is then designed to identify ground and non-ground patches within the digital elevation model before labeling their inner near-ground voxel centroids as GF-support nodes. The composite morphological filter is used to recognize non-ground areas with incomplete edge structures depicted in the input point cloud and to correct misclassified areas. Finally, a bidirectional <em>k</em>-dimensional tree search engine is built between the GF-support nodes and the input point cloud to separate ground and non-ground points. Experimental results show that DMF achieves ground filtering accuracy with an average F-score greater than 0.88, demonstrating robustness in generating digital terrain models across various test scenarios. Furthermore, the intermediate outputs of DMF enable instance segmentation of artificial objects in airborne point clouds. The code for DMF will be shared on GitHub (<span><span>https://github.com/wbx1727031/DMF</span><svg><path></path></svg></span>).</div></div>","PeriodicalId":100730,"journal":{"name":"ISPRS Open Journal of Photogrammetry and Remote Sensing","volume":"18 ","pages":"Article 100102"},"PeriodicalIF":0.0,"publicationDate":"2025-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145190142","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Drone imaging-based wall-to-wall processing pipelines for individual tree level inventory in boreal forest plots","authors":"Olli Nevalainen ,&nbsp;Niko Koivumäki ,&nbsp;Raquel Alves de Oliveira ,&nbsp;Teemu Hakala ,&nbsp;Roope Näsi ,&nbsp;Xinlian Liang ,&nbsp;Yunsheng Wang ,&nbsp;Juha Hyyppä ,&nbsp;Eija Honkavaara","doi":"10.1016/j.ophoto.2025.100099","DOIUrl":"10.1016/j.ophoto.2025.100099","url":null,"abstract":"<div><div>Precise individual tree data are essential for forest management, strategic planning, efficient commercial forestry, and accurate carbon stock assessments. In this study, a wall-to-wall drone-imaging-based forest inventory processing pipeline was developed and assessed. Different cameras and data analysis methods were assessed for individual tree detection and attribute estimation at the tree and plot levels. The experiment was conducted in Finland in six boreal forest study areas, with three major tree species: Scots pine <em>(Pinus sylvestris),</em> Norway spruce (<em>Picea abies</em>), and birch <em>(Betula pendula</em> and <em>Betula pubescens).</em> RGB and multispectral (MS) cameras provided single-sensor solutions for the forest inventory pipeline, whereas a hyperspectral (HS) camera was used in combination with the RGB camera to enhance species classification. High-quality RGB data performed better than MS data for tree detection and attribute estimation. The best tree detection rates were 56–84 % in areas with mostly dominant and co-dominant trees. The two evaluated tree detection methods (local maximum and segmentation) provided similar tree detection rates and tree attribute estimation accuracies. Tree level attributes were estimated with root mean square errors (RMSEs) of 0.97 m (5.1 %) for tree height, 3.1 cm (14 %) for diameter at breast height (DBH), 129.6 cm² (25 %) for the basal area, and 0.13 m³ (23 %) for the volume. The HS camera yielded the highest tree species classification performance, with maximum f-scores of 0.81 for RGB, 0.88 for MS, and 0.89 for combined HS + RGB data. At the plot level, RMSEs for stem density, basal area, and volume were 855.7 ha^-1 (74.6 %), 6.9 m² ha⁻¹ (24.2 %), and 48.6 m³ ha⁻¹ (17.6 %), respectively. This study was the first to assess entire inventory pipelines with a comprehensive camera setup and proved that low-cost RGB and MS cameras provide acceptable performance for tree inventories in boreal forests. These results can guide the implementation of low-cost forest inventory processes.</div></div>","PeriodicalId":100730,"journal":{"name":"ISPRS Open Journal of Photogrammetry and Remote Sensing","volume":"17 ","pages":"Article 100099"},"PeriodicalIF":0.0,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144903839","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A structured review and taxonomy of next-best-view strategies for 3D reconstruction","authors":"Bashar Alsadik ,&nbsp;Hussein Alwan Mahdi ,&nbsp;Nagham Amer Abdulateef","doi":"10.1016/j.ophoto.2025.100098","DOIUrl":"10.1016/j.ophoto.2025.100098","url":null,"abstract":"<div><div>Next-Best-View (NBV) strategies are a class of approaches that solve the important problem of selecting the best possible viewpoints of an autonomous robot sensor for effective and complete 3D scene reconstruction. NBV methodologies have developed significantly over the years from rule-based approaches to those driven from deep learning. Consequently, NBV strategies have become diverse and uncategorized which makes it difficult for researchers and practitioners to navigate or standardize the methods. Therefore, in this paper, a comprehensive review was conducted to separate NBV methods into five distinct strategies: rule-based, uncertainty-based, sampling-based, learning-based, and prediction-based approaches. It is aimed to give a structured understanding after systematically reviewing over 100 publications including outlining key methodologies, open-access tools, and respective applications. Each strategy is investigated with related research questions such as understanding the role of geometric heuristics in rule-based methods, identifying efficient sampling mechanisms for exploration, leveraging predictive models for optimization, addressing uncertainty in unknown environments, and applying learning-based techniques to enhance adaptability and performance. Some suggestions are made for making classifications explicit, thus helping pull together more organized frameworks and collaborations across disciplines. This work not only offers a comprehensive resource for beginners and expert researchers but also empowers readers to answer strategy-specific research questions, providing actionable insights into NBV planning trends and emerging perspectives.</div></div>","PeriodicalId":100730,"journal":{"name":"ISPRS Open Journal of Photogrammetry and Remote Sensing","volume":"17 ","pages":"Article 100098"},"PeriodicalIF":0.0,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144861224","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Asynchronous Lidar: Proof-of-concept simulation and demonstration tests","authors":"Craig L. Glennie ,&nbsp;Luyen K. Bui ,&nbsp;Francisco Haces-Garcia ,&nbsp;Derek D. Lichti","doi":"10.1016/j.ophoto.2025.100096","DOIUrl":"10.1016/j.ophoto.2025.100096","url":null,"abstract":"<div><div>This study proposes an asynchronous airborne lidar design in which the laser transmitter and detectors/receivers are disconnected and carried on separate platforms. This design is more advantageous than conventional synchronous lidar systems operating in monostatic mode because redundant lidar observations can be captured. First, proof-of-concept experiments are conducted based on Monte Carlo simulations assuming a transmitter is combined with different numbers of receivers. In this way, different receiver configurations, i.e., the locations of the transmitter and receivers relative to each other, are tested with both single beam (nadir and slant range) and multi beam transmitters. Networks with the transmitter, receivers, and ground point forming a plane result in very high dilution of precision corresponding to high ground point uncertainties, which are weak configurations and should be avoided. A laboratory demonstration of an asynchronous lidar system is also presented. The results from the lab demo validate the observations made by the simulation studies. Networks with three or four receivers appear to be a reasonable balance between the number of receivers used and the ground point uncertainties. Ground point uncertainties are also dependent on the transmitter and receiver flight altitudes; multi beam simulations of four-receiver networks with varying transmitter/receiver flight heights show that the horizontal uncertainties are almost completely dependent on the transmitter flight altitude, however, both flight altitudes affect the vertical uncertainty with the receiver flight altitude having a greater influence. The best configuration with the lowest uncertainties is obtained by maximizing the ratio of transmitter height to receiver height.</div></div>","PeriodicalId":100730,"journal":{"name":"ISPRS Open Journal of Photogrammetry and Remote Sensing","volume":"17 ","pages":"Article 100096"},"PeriodicalIF":0.0,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144829175","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A method for extracting water surface and hydrophytic vegetation from ICESat-2 data in wetlands","authors":"Rong Zhao ,&nbsp;Shijuan Gao ,&nbsp;Kun Zhang ,&nbsp;Defang Li ,&nbsp;Yi Li","doi":"10.1016/j.ophoto.2025.100097","DOIUrl":"10.1016/j.ophoto.2025.100097","url":null,"abstract":"<div><div>The Ice, Cloud, and Land Elevation Satellite-2 provides a great opportunity to measure water surface and hydrophytic vegetation in complex wetlands. Obtaining reliable signal photons from ICESat-2 data in wetlands is challenging because there are many types of noise photons, such as specular return photons, after-pulse photons, and noise photons caused by sunlight. In addition, the high photon density difference between the water and hydrophytic vegetation makes it difficult to find accurate hydrophytic vegetation photons. Therefore, this research aims to propose a method to obtain high-accuracy signal photons and classify water body photons and hydrophytic vegetation photons in complex wetlands. First, we introduced the modified elevation histogram statistics vector-based (MEHSV) method to filter out noise photons caused by sunlight. The MEHSV method was developed to retain sparse canopy photons. Therefore, the MEHSV method can retain sparse hydrophytic vegetation photons. Second, peak analysis of the elevation histogram statistics removed the specular return photons and after-pulse photons caused by the water surface. Finally, the manually labeled photons and reference water surface level data were used to assess the proposed method. The filtering results showed that the F value of the proposed method achieved 0.99. Compared with other reference methods, the proposed method both preserved hydrophytic vegetation photons being misrecognized and removed all types of noise photons effectively. The water photons and hydrophytic vegetation photons were distinguished accurately. Additionally, the accuracy of water surface level (R² = 0.97, and RMSE = 0.84 m) witnessed the good performance of the proposed method.</div></div>","PeriodicalId":100730,"journal":{"name":"ISPRS Open Journal of Photogrammetry and Remote Sensing","volume":"17 ","pages":"Article 100097"},"PeriodicalIF":0.0,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144903838","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"From gaps to granularity: CRPAG-DSHAT based multi-modal deep learning framework for DEM void repair and super-resolution reconstruction in Himalayas","authors":"Sayantan Mandal,&nbsp;Ashis Kumar Saha","doi":"10.1016/j.ophoto.2025.100101","DOIUrl":"10.1016/j.ophoto.2025.100101","url":null,"abstract":"<div><div>Digital Elevation Models (DEMs) are essential for terrain characterization and environmental modeling, yet their utility is limited by data voids and coarse resolution, especially in complex mountainous regions of Himalayas. To address these challenges, we propose a novel dual-stage deep learning pipeline that unifies void filling and super-resolution into a cohesive framework, leveraging both topographic fidelity and spectral texture. First, the <strong>Conditional Residual Pyramid Attentional Generator (CRPAG)</strong> a hybrid model that integrates multi-scale DEM features with Sentinel-2 red band reflectance (∼665 nm) using an <strong>Improved Channel Attention Module</strong> (ICAM), <strong>Residual Pyramid Attention Block</strong> (TFG_RPAB), and a dual-encoder design. This allows CRPAG to prioritize structural fidelity (RMSE 9.1–28.9 m) while reconstructing missing terrain features (Mean Absolute Error MAE 1.9–8.1 m). This void-filled, high-resolution DEM then supervise the training of <strong>Dual-Stream Hierarchical Attention Transformer (DS-HAT)</strong>, which performs super-resolution on globally available low-resolution DEMs (ALOS PALSAR), guided by pixel-wise height attention and texture-aware mechanisms. Compared to benchmark models such as MCU-Net-EDF and conventional U-Nets, our integrated system shows improvements in elevation accuracy (RMSE ↓, P95 = 9.2 m), spatial consistency (Moran's I ↑), and structural similarity (SSIM ↑), particularly across high-curvature and spectrally ambiguous regions. Besides, Ablation studies confirm the complementary applications of topographic variables in mitigating oversmoothing and enhancing terrain realism. This dual-stage strategy not only enhances DEM fidelity but also provides a scalable framework for improving DEM quality. Through this multi-modal fusion, this work transforms topographic knowledge into computable framework, advancing DEM applicability in hydrological modeling, detection mechanisms and disaster prediction.</div></div>","PeriodicalId":100730,"journal":{"name":"ISPRS Open Journal of Photogrammetry and Remote Sensing","volume":"17 ","pages":"Article 100101"},"PeriodicalIF":0.0,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145009967","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"FeatureGS: Eigenvalue-feature optimization in 3D Gaussian Splatting for geometrically accurate and artifact-reduced reconstruction","authors":"Miriam Jäger,&nbsp;Markus Hillemann,&nbsp;Boris Jutzi","doi":"10.1016/j.ophoto.2025.100100","DOIUrl":"10.1016/j.ophoto.2025.100100","url":null,"abstract":"<div><div>3D Gaussian Splatting (3DGS) has emerged as a powerful approach for 3D scene reconstruction using 3D Gaussians. However, neither the centers nor surfaces of the Gaussians are accurately aligned to the object surface, complicating their direct use in point cloud and mesh reconstruction. Additionally, 3DGS typically produces floater artifacts, increasing the number of Gaussians and storage requirements. To address these issues, we present FeatureGS, which incorporates an additional geometric loss term based on an eigenvalue-derived 3D shape feature into the optimization process of 3DGS. The goal is to improve geometric accuracy and enhance properties of planar surfaces with reduced structural entropy in local 3D neighborhoods, typically given in man-made environments. We present four alternative formulations for the geometric loss term based on ‘planarity’ of Gaussians, as well as ‘planarity’, ‘omnivariance’, and ‘eigenentropy’ of Gaussian neighborhoods. On the small-scale DTU benchmark with man-made scenes, FeatureGS achieves a 20% improvement in geometric accuracy, suppresses floater artifacts by 90%, and reduces the number of Gaussians by 95%. FeatureGS proves to be a strong method for geometrically accurate, artifact-reduced and memory-efficient 3D scene reconstruction, enabling the direct use of Gaussian centers for geometric representation.</div></div>","PeriodicalId":100730,"journal":{"name":"ISPRS Open Journal of Photogrammetry and Remote Sensing","volume":"17 ","pages":"Article 100100"},"PeriodicalIF":0.0,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144988574","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Permanent terrestrial laser scanning for near-continuous environmental observations: Systems, methods, challenges and applications","authors":"Roderik Lindenbergh ,&nbsp;Katharina Anders ,&nbsp;Mariana Campos ,&nbsp;Daniel Czerwonka-Schröder ,&nbsp;Bernhard Höfle ,&nbsp;Mieke Kuschnerus ,&nbsp;Eetu Puttonen ,&nbsp;Rainer Prinz ,&nbsp;Martin Rutzinger ,&nbsp;Annelies Voordendag ,&nbsp;Sander Vos","doi":"10.1016/j.ophoto.2025.100094","DOIUrl":"10.1016/j.ophoto.2025.100094","url":null,"abstract":"<div><div>Many topographic scenes exhibit complex dynamic behavior that is difficult to map, quantify, predict and understand. A terrestrial laser scanner fixed on a permanent position can be used to monitor such scenes in an automated way with centimeter to decimeter quality at ranges of up to several kilometers. Laser scanners are active sensors, and are therefore able to continue operation during night. Their independence from texture conditions ensures that in principle they provide stable range measurements for varying surface conditions. Recent years have seen a strong increase in the employment of such systems for different scientific applications in geosciences, environmental and ecological sciences, including forestry, glaciology, and geomorphology. At the same time, this employment resulted in a new type of 4D topographic data sets (3D point clouds + time) with a significant temporal dimension, as systems are now able to acquire thousands of consecutive epochs in a row. Extracting information from these 4D data sets turns out to be challenging, first, because of insufficient knowledge on error budget and correlations, and, second, because of lack of algorithms, benchmarks, and best-practice workflows. This paper provides an overview of different 4D systems for near-continuous laser scanning, and discusses systematic challenges including instability of the sensor system, meteorological and atmospheric influences, and data alignment, before discussing recently developed methods and scientific software for extracting and parameterizing changes from 4D topographic data sets, in connection to the different applications.</div></div>","PeriodicalId":100730,"journal":{"name":"ISPRS Open Journal of Photogrammetry and Remote Sensing","volume":"17 ","pages":"Article 100094"},"PeriodicalIF":0.0,"publicationDate":"2025-07-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144604308","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Evaluating the role of training data origin for country-scale cropland mapping in data-scarce regions: A case study of Nigeria","authors":"Joaquin Gajardo ,&nbsp;Michele Volpi ,&nbsp;Daniel Onwude ,&nbsp;Thijs Defraeye","doi":"10.1016/j.ophoto.2025.100091","DOIUrl":"10.1016/j.ophoto.2025.100091","url":null,"abstract":"<div><div>Cropland maps are essential for remote sensing-based agricultural monitoring, providing timely insights about agricultural development without requiring extensive field surveys. While machine learning enables large-scale mapping, it relies on geo-referenced ground-truth data, which is time-consuming to collect, motivating efforts to integrate global datasets for mapping in data-scarce regions. A key challenge is understanding how the quantity, quality, and proximity of the training data to the target region influences model performance in regions with limited local ground truth. To address this, we evaluate the impact of combining global and local datasets for cropland mapping in Nigeria at 10 m resolution. We manually labelled 1,827 data points evenly distributed across Nigeria and leveraged the crowd-sourced Geowiki dataset, evaluating three subsets of it: Nigeria, Nigeria + neighbouring countries, and worldwide. Using Google Earth Engine (GEE), we extracted multi-source time series data from Sentinel-1, Sentinel-2, ERA5 climate, and a digital elevation model (DEM) and compared Random Forest (RF) classifiers with Long Short-Term Memory (LSTM) networks, including a lightweight multi-task learning variant (multi-headed LSTM), previously applied to cropland mapping in other regions. Our findings highlight the importance of local training data, which consistently improved performance, with accuracy gains up to 0.246 (RF) and 0.178 (LSTM). Models trained on Nigeria-only or regional datasets outperformed those trained on global data, except for the multi-headed LSTM, which uniquely benefited from global samples when local data was unavailable. A sensitivity analysis revealed that Sentinel-1, climate, and topographic data were particularly important, as their removal reduced accuracy by up to 0.154 and F1-score by 0.593. Handling class imbalance was also critical, with weighted loss functions improving accuracy by up to 0.071 for the single-headed LSTM. Our best-performing model, a single-headed LSTM trained on Nigeria-only data, achieved an F1-score of 0.814 and accuracy of 0.842, performing competitively with the best global land cover product and showing strong recall performance, a metric highly-relevant for food security applications. These results underscore the value of regionally focused training data, proper class imbalance handling, and multi-modal feature integration for improving cropland mapping in data-scarce regions. We release our data, source code, output maps, and an interactive GEE web application to facilitate further research.</div></div>","PeriodicalId":100730,"journal":{"name":"ISPRS Open Journal of Photogrammetry and Remote Sensing","volume":"17 ","pages":"Article 100091"},"PeriodicalIF":0.0,"publicationDate":"2025-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144596268","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Efficient tree mapping through deep distance transform (DDT) learning","authors":"Jan Schindler ,&nbsp;Ziyi Sun ,&nbsp;Bing Xue ,&nbsp;Mengjie Zhang","doi":"10.1016/j.ophoto.2025.100095","DOIUrl":"10.1016/j.ophoto.2025.100095","url":null,"abstract":"<div><div>Trees provide essential ecosystem services in urban areas, rural landscapes and forests. Individual tree information can inform forest and risk modelling, health studies and decision-making in public and non-governmental sectors. The increase in available remote sensing data and advances in automated object detection makes it feasible to map trees over large areas in unprecedented detail. Deep learning-based instance segmentation methods have thereby become the state-of-the-art in tree crown delineations tasks from aerial ortho-photography. Many of these methods are based on one- and two-stage detector frameworks such as Mask-RCNN and YOLO, which were developed focussing on speed and accuracy against common benchmark datasets. Another class of object detectors is based on encoder-decoder networks such as UNet which offer easy integration into existing workflows and high accuracy even in complex forest scenes in regional and national tree studies. While previous methods had to combine multi-model and multi-task outputs to create decision surfaces, we developed an efficient UNet-based modelling approach which focusses solely on learning the distance transforms of tree objects as cost surface for watershed segmentation. Our algorithm achieves superior instance segmentation across native forest, rural and urban environments in Aotearoa New Zealand, with an overall F1 score of 0.53 — 0.18 for small, 0.45 for medium and 0.67 for large crowns — surpassing previous approaches while decreasing modelling complexity, enabling fast and large-scale tree mapping.</div></div>","PeriodicalId":100730,"journal":{"name":"ISPRS Open Journal of Photogrammetry and Remote Sensing","volume":"17 ","pages":"Article 100095"},"PeriodicalIF":0.0,"publicationDate":"2025-06-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144595538","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}